^{page 27 note 1} I have made a textual suggestion (‘An Early Revision of the Gospel of Mark’, J.B.L. LXXVIII (1959), 215–27) which would eliminate most of the ‘minor agreements’ between Matthew and Luke.Google ScholarMgr, de Solages, A Greek Synopsis of the Gospels (Leiden, 1959), pp. 1055–65, now gives the fullest list of such agreements.Google Scholar For the text of Mark known to Matthew, and Luke, , cf., A. F. J. Klijn, ‘A Survey of the Researches into the Western Text of the Gospels and Acts’, II, Novum Testamentum, III (1959), 162 note 2.Google Scholar

^{page 27 note 2} Wilhelm, Bussmann, Synoptische Studien,: II, Zur Redenquelle (Halle, 1929), pp. 137–49.Google Scholar

^{page 27 note 3} Wilfred, L. Knox, The Sources of the Synoptic Gospels: II, St Luke and St Matthew (Cambridge, 1957), ed. Chadwick, H..Google Scholar

^{page 27 note 4} Joachim, Jeremias, ‘Zur Hypothese einer schriftlichen Logienquelle Q’, Z.N.T.W. XXIX (1930), 147–9.Google Scholar

^{page 27 note 5} A survey of possible explanations in Vincent, Taylor, ‘The Elusive Q’, E.T. XLVI (1934–1935), 68–74. I sharpen his proposal (2): Matthew and Luke used different ‘recensions’ of Q.Google Scholar I have not seen the recension-theory in Patten's, C. S.Sources of the Synoptic Gospels (1915).Google Scholar

^{page 28 note 1} First seen by WelLhausen and Holtzmann: cf., A. H. M'Neile, The Gospel According to St Matthew (London, 1928), p. 129.Google Scholar M'Neile himself thought that Matt., ix. 27–34 was added ‘to complete a triplet’ with ix. 18–26;Google Scholar others (e.g. Kilpatrick, G. D., The Origins of the Gospel According to St Matthew (corr. ed., Oxford, 1950, p. 90) that Matthew ‘may have wished to make up a list of ten miracles’. These views rather cancel each other out.Google Scholar

^{page 28 note 2} Luke introduces the Baptist's question by the raising of the widow's son (vii. 11–17) Out of the same motive.

^{page 28 note 3} Cf. in fact άπτοματ (Mark, vii. 33 and viii.22;Google ScholarMatt., ix. 29);Google Scholar ‘open’ (Mark, vii. 34;Google ScholarMatt., ix. 30);Google Scholar ‘they carry’ (Mark, vii. 32;Google Scholar Matt. ix. 32 xii. 22); the command of secrecy (Mark, vii. 36; viii. 26;Google ScholarMatt., ix. 30).Google Scholar

^{page 28 note 4} Was not Matthew lucky to find in Mark just the narratives he needed? Perhaps those narratives were shaped by the Q-source of Matt., xi. 5 and the Isaianic prophecies it summarizes.Google ScholarMark, vii. 32–5Google Scholar strongly echoes Isa., xxxv. 5–6 LXX; cf. especially the rare word μογιλάλος.Google Scholar

^{page 28 note 5} Except that passages suspected of being late additions to Q (the Temptations; Jesus' thanks giving, Luke, x. 21–2; ‘Jerusalem’, xiii. 34–5) have large agreement in Greek. Perhaps in fact they are Greek compositions, so that no Aramaic ever existed to be a disturbing factor behind Matthew.Google Scholar

^{page 29 note 1} Kilpatrick, , op. cit. pp. 37–58.Google Scholar

^{page 29 note 2} Jeremias, J., The Parables of Jesus (tr. Hooke, S. H., London, 1954), pp. 63–8.Google Scholar

^{page 29 note 3} Possibly something similar stood in ‘Q^mt’ at Matt., xiii. 49;Google Scholarcf., note 4, p. 36 below. But the phrase seems derived from συντελεīσθαι,Google Scholar (Mark, xiii. 4) under the influence of είς συντέλειαν έμερνGoogle Scholar(Daniel, xii. 13, LXX and Theodotion), so that Matthew goes from Genesis (i. I) to Consummation (xxviii. 20).Google Scholar

^{page 29 note 4} Summary of the problem: Grant, F. C., ‘Further Thoughts on the M-Hypothesis’, E. T. XLVI (1934–1935), 438–45.Google Scholar

^{page 29 note 5} Nearly following Kilpatrick, (op. cit. p. 24);Google Scholarcf., Alfred M. Perry, ‘The Framework of the Sermon on the Mount’, J.B.L. LIV (1935), 103–15.Google Scholar

^{page 30 note 1} Filson, F. V., ‘Broken Patterns in the Gospel of Matthew’, J.B.L. LXXV (1956), 227–31.Google Scholar

^{page 30 note 2} Kilpatrick, G. D., ‘The Gentile Mission in Mark and Mark 13. 9–11’, Studies in the Gospels: Essays in Memory of R. H. Lightfoot (ed. Nineham, D. E., Oxford, 1955), pp. 145–58.Google Scholar

^{page 30 note 3} Its original is Matt., x. 5–6 (Q-context) which, if genuine, must have meant ‘Go [not to the Pharisees, but] to the lost sheep of the house of Israel’.Google Scholar

^{page 30 note 4} ‘Do not even the publicans/Gentiles do the same?’ (Matt., v. 46–7) is a compliment; but apparently ‘Gentiles’ is editorial even here.Google Scholar For combining Matt., v. 46–7Google Scholar and Luke, vi. 32–4Google Scholar one concludes that Greek Q had ‘publicans/sinners’ here as at Matt., xi. 19Google ScholarLuke, vii. 34Google Scholar (Q) and Mark, ii. 15–16Google Scholar (used at Luke, xv. 1).Google Scholar But ‘sinners’ is too vague for a good parallel; we seem to have a better Greek version of the same Aramaic original at Matt., xxi. 31–2Google Scholar (Q?—cf., Luke vii. 29–30), ‘publicans and harlots (πόρναι)’.Google ScholarJustin, , Apology xv, 9–10Google Scholar a version of Matt., v. 46–7Google ScholarLuke, vi.32–4)Google Scholar has ‘Even the πόρνοι do this…even the publicans do this’. Then the original meaning in Matt., v. 46–7 (Q); xi.19(Q); xxi. 31–2;Google ScholarMark, ii. 15–16 was ‘publicans/prostitutes (male or female)’.Google Scholar So in the formula of excommunication, Matt., xviii. 17 has ‘Gentile and publican’,Google Scholar I Cor., v. 11 πόρνος πλεονέκτης which perhaps between them reflect the same parallelism.Google Scholar The origin of ‘Gentile’ at Matt., vi. 5–7 (with ‘hypocrites’) and x. 5 (with ‘Samaritans’) is less obvious.Google Scholar

^{page 31 note 1} Streeter, B. H., The Four Gospels (rev. ed., London, 1930), pp. 238–65.Google Scholar

^{page 31 note 2} Manson, T. W., The Sayings of Jesus (= Part II of The Mission and Message of Jesus, ed. Major, H. D., 1937), London, 1949, pp. 21–6.Google Scholar

^{page 31 note 3} Op. cit. p. 35.Google Scholar

^{page 31 note 4} So Bussmann looked for a series of stages in the Marcan narrative-tradition rather than in the Q sayings-tradition, whose verbal form is much more strongly fixed.

^{page 32 note 1} Except the widow's mite (Mark, xii. 41–4),Google Scholar dropped because Matthew wanted to make one sermon out of the anti-Pharisaic polemic (Mark, xii. 38–40) and the apocalypse (Mark xiii).Google ScholarCf., note 4, p. 35.Google Scholar

^{page 32 note 2} Parker, P. (The Gospel Before Mark, Chicago, 1953) holds that all classes of Matthew's special materials were once united (not with Q but) with Mark in a document ‘K’. But his reconstruction of K (pp. 188–235) includes almost exactly Manson's M-document, and falls under the same difficulty of extensive overlap with Q.Google Scholar

^{page 32 note 3} Matt., XII. 5–7 is not ‘M’ but oral tradition that has grown up around Mark.Google ScholarMatt., xv. 13 (‘Every plant that my Father has not planted’) adheres more closely to its Q-context than its Marcan.Google ScholarMatt., xix. 10–12,Google Scholar the eunuchs, was perhaps comment on the divorce-saying in its Q-forrn (Matt., v. 32)Google Scholar rather than in its Marcan (Matt., xix. 9).Google ScholarJustin, , Apology xv, 1–4Google Scholar gives in succession Matt., v. 28, 29, 32; xix, 11–12, all in variant form: perhaps they came to him united.Google Scholar Ignatius, who elsewhere echoes only Q^mt (below), knows matt.XV. 13 (Trall. xi. 1;Google ScholarPhilad. iii. 1) and xix.12Google Scholar (Smyrn. vi.1.Google Scholar

^{page 33 note 1} ‘The Book of Sayings used by the Editor of the First Gospel’, Studies in the Synoptic Problem (ed. Sanday, W., Oxford, 1911), pp. 235–86.Google Scholar

^{page 34 note 1} But Matt., xi. 28–30 cannot all have come from Aramaic or all stood in Q^mt.Google Scholar It is partly accident that it goes so well into Syriac (sy^s) and Aramaic (Black, M., An Aramaic Approach to the Gospels and Acts (2nd ed., Oxford, 1954), p. 140).Google Scholar For it echoes Sirach, li. 23–7,Google Scholar where the Hebrew (ed. Israel, Lévi, Leiden, 1904) does not give a form of the root Þßℶ for άνάπαυσινGoogle Scholar (Sirach, li. 27)Google Scholar on which the word-play in the Aramaic versions of Matt., xi.29 depends. Furthermore, Matthew is using Mark here: with △ευ˜τε……άναπαύσω;Google Scholar (Matt., xi. 28) cf. δευ˜τε…άναπαύσασθεGoogle Scholar (Mark, vi. 31, omitted by Matthew in context).Google Scholar

^{page 34 note 2} Krister, Stendahl, The School of St Matthew (Uppsala, 1954).Google Scholar

^{page 35 note 1} The lament over Jerusalem (Matt., xxiii. 37–9 =Google ScholarLuke, xiii. 34–5) has been thought the original Q-continuation of the quotation from the ‘wisdom of God’Google Scholar (Luke, xi. 49–51 =Google ScholarMatt., xxiii. 34–6).Google Scholar But the saying (after the Lucan editorial addition Luke, xiii. 31–3 is dropped)Google Scholar continues exactly the thought of Luke, xiii. 24–30,Google Scholar the rejection of Israel in favour of the Gentiles. It was perhaps a later addition slightly out of place, intended (with its first person) for the eschatological ‘house-holder’ of Luke, xiii. 25.Google Scholar

^{page 35 note 2} Butler, B. C., The Originality of St Matthew (Cambridge, 1951), p. 45.Google Scholar

^{page 35 note 3} Cf. the way(s) of light and darkness in the Qumran, Manual of Discipline iii. 20–1Google Scholar and Barnabas xviii ff.;Google Scholar and the ‘way of life’ and ‘way of death’ in Didache i–v.Google Scholar

^{page 35 note 4} The story of the strange exorcist (Mark, ix. 38–9)Google Scholar is related to Matt., vii. 22 cf. ‘in thy name’, δαιμόνια έκβάλλειν, δύναμιν (δυνάμεις) ποιειν.Google Scholar Matthew omits Mark, ix. 38–40Google Scholar in context as if he were using it at Matt., vii. 22,Google Scholar but this may be coincidence: Justin, , Apology xvi, 11Google Scholar (below) suggests that the edited version Matt., vii. 22 already stood in Q^mt; perhaps Mark already reflects Q^mt.Google Scholar On the other hand, Luke, x. 17–19,Google Scholar the report of the Seventy, seems clearly to use Mark, ix. 38–9—and therefore did not come to Luke from ‘L’.Google Scholar

^{page 36 note 1} The Q-materials Out of order here which break M-connexions are plainly late additions. Several must at one stage in the making of Q^mt have stood in sermon IV. For Mark and Luke agree against Matthew in putting those materials in their Church-discipline sections (Mark, ix. 41-xl. 25;Google ScholarLuke, xiv. 34-xvii. 6):Google Scholar on salt, Matt., v. 13Google Scholar (Mark, ix. 50;Google ScholarLuke, xiv. 34–5);Google Scholar the permanence of the Law, v. 17–18 (Mark, xiii. 31;Google ScholarLuke, xvi. 17);Google Scholar the offending member, v. 29–30 (Mark, ix. 43–7);Google Scholar divorce, v. 32 (Mark, x. 11–12;Google ScholarLuke, xvi. 18);Google Scholar on anxiety, vi. 19–21, 25–34 (Mark, x. 21;Google ScholarLuke, xii. 22–34).Google Scholar

^{page 36 note 2} ‘There is no one who has given up…mother or father…for my sake’ (Mark, x. 29) seems to reflect the edited version of Q in ‘he that loves father or mother more than me’Google Scholar (Matt., x. 37).Google Scholar Perhaps then the missing refrain of sermon II in Q^mt is preserved by ‘the first shall be last’ (Mark, x. 31Google Scholar = Luke, xiii. 30).Google Scholar

^{page 36 note 3} ‘Western’ additions to Matthew sometimes seem to be a reading of Q^mt, suppressed by the final editor of Matthew, but restored in the later history of the text. So [Matt., ] xvi. 2–3Google Scholar is the missing Q^mt-version of Luke, xii. 54–6 (apparently Q);Google Scholar the Western text of Matt., v. 44Google Scholar restores the full reading of Luke, vi. 27–8;Google Scholar [Matt., ] xx. 28aGoogle Scholar is the missing version of Luke, xiv. 7–10 (note 5, below).Google Scholar Cf. further the variant readings at Matt., Vii 21–2; xviii.11, 20; xx. 16; xxiii. 13 (sy^s).Google Scholar

^{page 36 note 4} Unless the three parables Matt., xiii. 44–8 are original Q, omitted by Luke, they are Q^mt supplements homogeneous in style to xiii. 31–3. Then ‘weeping and gnashing’ (xiii. 50), as an editorial formula broken by the final editor, must have been appended to the parables in Q^mt. This seems to verify the eschatological application xiii. 49–50 for Q^mt also, apart from έν τῇ συντελείᾳ τού αίνος—unless the final editor found this in Q^mt and adapted it (note 3, p. 29).Google Scholar The interpretation of the parable of the tares (xiii. 36–43) combines two unrelated motifs: Mark's method of parable-interpretation (Mark, iv. 13–20)Google Scholar and the eschatology of Matt., xiii. 49–50. The parable of the Tares itself (xiii. 24–30) may contain a written nucleus similar to the Net (xiii. 47–8); but in any case Matthew has worked it up very freely with phrases from Mark's seed-parable (iv. 26–9). There are difficulties in this reconstruction; but on any analysis Matt. xiii is an elaborate piece of combination and rewriting.Google Scholar

^{page 36 note 5} The banquet-parable Luke, xiv. 7–10 is Q.Google Scholar Its gnomic conclusion (the high and the low) is given by both Matt., xxiii. 12 and Luke xiv. 11 in the Context of ‘Jerusalem’Google Scholar (Matt., xxiii. 37–9Google Scholar = Luke, xiii. 34–5).Google Scholar Furthermore, Matthew knew that conclusion in the parable-context; for he has inserted it against the rubric πρω;τοκλισίας (Mark, xii. 39Google Scholar = Matt., xxiii. 6)Google Scholar in the anti-Pharisaic polemic, while suppressing the reference (πρω;τοκισίαν Luke, xiv. 8) which would have made his procedure intelligible. Cf. note 3, above.Google Scholar

^{page 36 note 6} Matt., xxi. 28–31Google Scholar perhaps reflects Luke, xii 47–8Google Scholar (which will then be Q): cf. ‘do the will’, ‘lord’, baptism, δοκεīν (Luke, xii. 51).Google Scholar Furthermore, Matthew omits ‘I have a baptism to be baptized, and how I am straitened έω;ς ότον τελεσθῇ’ (Luke, xii. 50, presumably Q);Google Scholar and adds πληρσαι πσαν δικαιοσύνην to the Baptism-narrative (Matt., iii. 15). Are not these two halves of one operation? It will remove the obscure metaphor ‘baptism = death’, and explain the awkward fact of Jesus' baptism by John.Google Scholar

^{page 36 note 7} The passage Matt., xxiv. 10–13,30–1 must once have existed uncombined with Marcan materials;Google Scholar for it is precisely the Synoptic matter echoed in the apocalypse of Didache xvi.3–8 (below).Google Scholar

^{page 37 note 1} Cf., Justin, Dialogue LXXVI, 4–5: 'They shall come from the east and the west and lie down with Abraham and Isaac and Jacob in the kingdom of heaven; but the sons of the kingdom shall be cast into outer darknessGoogle Scholar (Matt., viii. 11–12a).Google Scholar …Many will say to me in that day, “Lord, Lord, have we not in thy name eaten and drunk and prophesied and cast out demons?” And I shall say to them, “Depart from me” [a different combination of Matt., vii. 22–3Google Scholar and Luke, xiii. 26].…Google Scholar Depart to the Outer darkness, which my Father has prepared for Satan and his angels (cf., Matt. xxv. 41cod. D).’Google Scholar

^{page 37 note 2} Also found at Justin, , Dialogue xxxv, 3Google Scholar along with the canonical text, and echoed at Didache xvi. 3 and Didascalia vi.13Google Scholar (Connolly, , p. 210)—all three places in the context of the Synoptic Apocalypse.Google Scholar

^{page 37 note 3} The fundamental study is still Bousset, W., Die Evangeliencitate Justins des Mätyrers… (Göttingen, 1891).Google Scholar

^{page 37 note 4} Connolly, R. H., Didascalia Apostolorum (Oxford, 1929).Google Scholar For example, ‘there shall be schisms and heresies’ (Justin, , Dialogue xxxv, 3; Didascalia vi.5Google Scholar (Connolly, , p. 198));Google Scholar ‘there shall arise many false Christs…and they shall lead astray many’ (Justin, , Dialogue xxxv, 3;Google Scholarcf., Didascalia vi. 13Google Scholar (Connolly, , p. 210)Google Scholar and contrast Matt., xxiv. 11, 24;Google Scholar ‘which my Father has prepared for Satan and his angels’ (Justin, , Dialogue LXXVI, 5 (above);Google ScholarDidascalia vi. 23Google Scholar (Connolly, , p. 256)).Google Scholar

^{page 38 note 1} édouard, Massaux, Influence de l'évangile de saint Matthieu sur la littérature chrétienne avant saint Irénée(Louvain, 1950): Universitas catholica lovanensis, Dissertationes…conscriptae, series II, tom. 42.Google Scholar

^{page 38 note 2} Helmut, Köster, Synoptische überlieferung bei den apostolischen Vätern (Berlin, 1957):Google ScholarT.U. 65. K. also promises us a study of Justin.Google Scholar

^{page 38 note 3} The Oxford Society of Historical Theology, The New Testament in the Apostolic Fathers (Oxford, 1905).Google Scholar

^{page 38 note 4} Guillaumont, A.et al., The Gospel According to Thomas (Leiden, 1959).Google Scholar

^{page 39 note 1} Mark, iv. 24Google Scholar (= Matt., vii. 2), ‘with what measure’ may be intended to mean that an interpreter of parables is rewarded. The absence of the original Q-Sermon from Mark, though explicable by the stress of persecution (as in Revelation), is a serious distortion.Google Scholar The Epistles are better balanced: Rom, . xii. 14;Google Scholar I Cor., iv. 12–13;Google Scholar I Pet., ii. 20; iii. 9; Epistle of Polycarp ii. 3.Google Scholar

^{page 40 note 1} μακάριόν έστιν μλλον διδόναι λαμβάνειν (Acts xx. 35) may also be a Beatitude made out of Q. The Ephesian sermon is authenticated as the only one standing in a ‘we’-section. The citation-formula (‘remembering the words of the Lord Jesus, that he said’) introduces Q at Acts xi. 16; I Clement, xiii. 1; xlvi. 7; Epistle of Polycarp ii. 3;Google ScholarJohn, xv. 19.Google Scholar May not Acts xx. 35 and Matt., x. 8Google Scholar δώπεάν έλάβετε, δω;ρεάν δότε (Q^mt) be related? Matthew goes on ‘take not gold or silver’ (x. 9) and Paul has just said, ‘I coveted no man’s silver or gold’ (Acts xx. 33). Further variants are given by D in Acts, Didache i. 5, I Clement, ii.1, etc.Google Scholar (cf., Köster, pp.231–2).Google Scholar

^{page 40 note 2} Col., iii. 12Google Scholar (= Eph., iv. 32 above) has σπλάγχνα οίτιρμου˜, χρηστότντα.Google Scholar

^{page 41 note 1} Glover, R., ‘The Didache's Quotations and the Synoptic Gospels’, N.T.S. v (1958–1959), 12–29.Google Scholar After the text was in type I saw Butler, B. C., ‘The Literary Relations of Didache, Ch. XVI’ (J. T.S. XI (n.s.), 1960, 265–83).Google Scholar Butler holds that Didache xvi is dependent on sources indistinguishable from Luke and Matthew, respectively; I suggest, on other grounds, that those sources were in fact united. The principal problem is the relation between ‘Those who endure…shall be saved’ (Didache xvi. 5)Google Scholar and ‘He who endures…shall be saved’ (Matt., xxiv. 13Google Scholar = Mark, xiii. 13).Google Scholar

^{page 41 note 2} Goodspeed, E. J., ‘The Didache, Barnabas, and the Doctrina’, Anglican Theological Rev. XXVII (1945), 228–47.Google Scholar

^{page 41 note 3} Manson, (op. cit. p. 23, note) finds the same structure underlying Q, his ‘M’, and the Didache.Google Scholar

^{page 42 note 1} (1) Hermas, , Sim. v, v, 2Google Scholar ό άγρόσ ό κόσμος ούτός έστιν must reflect Matt., xiii. 38 ό δέ άγρός έστιν ό κόσμος, the parable-interpretation due to the final editor.Google Scholar (2) Justin, , Dialogue XLIX, 5 quotes the editorial addition ‘Then the disciples understood that he was speaking to them about John the Baptist’Google Scholar (Matt., xvii. 13).Google Scholar (3) Heracleon, , frag. 35Google Scholar (Völker, W., Quellen zur Geschichte der christlichen Gnosis, Tübingen, 1932, p. 79Google Scholar = Origen, Commentary on John xiii. 49) ‘it is the Son of Man who sows’Google Scholar explains John, iv. 37 by Matt. xiii. 37 (cf. Hermas above).Google Scholar (4) Ptolemaeus, , Letter to Flora IV, 4, quotes Jesus'Google Scholar discourse on marriage in the form of Matt., xix. 8Google Scholar against Mark, x. 5;Google Scholar at IV, 11–13 he quotes text and order of Matt., xv. 4–7Google Scholar against Mark, vii. 6–13.Google Scholar